Audio temporal grounding on AudioGrounding
90.1R1@.3SpotSound-A
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SpotSound-AModel Category=Open-Source Models, Integration Backbone=Audio Flamingo 32026.04 | 90.1 | 74.8 | 70.3 | |
| SpotSound-QModel Category=Open-Source Models, Integration Backbone=Qwen2 Audio2026.04 | 87.2 | 66.1 | 67.8 | |
| TimeAudioModel Category=Open-Source Models2026.04 | 83.3 | 68.7 | 67.4 | |
| WTATGModel Category=Non-LLM Models2026.04 | 72.5 | 53.7 | 51.4 | |
| Audio Flamingo 3Model Category=Open-Source Models2026.04 | 66.9 | 42 | 47.5 | |
| AM-DETRModel Category=Non-LLM Models2026.04 | 52.5 | 15.6 | 30.2 | |
| Gemini-2.5-FlashModel Category=Proprietary Models2026.04 | 51.8 | 36.2 | 37.1 | |
| Qwen2-AudioModel Category=Open-Source Models2026.04 | 50.3 | 29.3 | 37 | |
| Gemini-2.5-ProModel Category=Proprietary Models2026.04 | 45.3 | 31.7 | 33.5 | |
| Kimi-AudioModel Category=Open-Source Models2026.04 | 4.6 | 2.8 | 4.9 |